JPH0447267B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION Field of the Invention The present invention relates to an apparatus for measuring the flight speed of a helicopter using periodic and simultaneous maneuver position signals and attitude angles for pitch and roll. PRIOR ART AND THE PROBLEM TO BE SOLVED BY THE INVENTION Due to the widespread use of modern helicopters, accurate knowledge of the current flight conditions, in particular the flight speed, is increasingly required. However, it has been recognized that flight speed detection is difficult in helicopters. Since the flight range includes extremely slow flight up to hovering and even backward flight, it is not possible to use the measurement principle of differential pressure measurement, which is actually used in airplanes, as the basis for speed measurement. Can not. Furthermore, in addition to the slow flight range, helicopters can also produce very large overall angles of attack and sideslip, so that fixed speeds, such as those used in winged vehicles, cannot be controlled. Sensors are not enough. Indeed, in helicopters it is almost impossible to find suitable locations for measuring sensors in a relatively large flight area, since all components are influenced more or less significantly by the main rotor and tail rotor beams. Therefore, it is prohibited to install conventional angle-of-attack and sideslip angle-measuring devices. At the same time speed measurements have proven difficult, but the great interest in flight speed measurement devices for helicopters has already led to a number of different measurement methods being proposed, as described below. Currently used sensor devices as well as indirect, so-called analytical methods that are under development meet the above requirements for accuracy, but they are either very costly or can only be used in certain flight regions, and the device It is hoped that this will be improved. For the reasons already mentioned, devices have been developed for measuring the flight speed specifically for slow flight in helicopters, in which case it is possible to distinguish between relatively direct and purely indirect measurement methods. Direct methods include, for example, LORAS (Low
Range Airspeed System). The sensor is
It consists of two bench lily tubes driven at 720 rpm by a motor, mounted at the end of a rotatably mounted arm and connected to a differential pressure generator. A typical installation location is above the rotor head. The difference in the negative pressure of the two ventilator tubes results in a sinusoidal signal in the airflow that runs through the device perpendicular to the axis of rotation during rotation. The maximum amplitude of this signal is proportional to the translation speed, and the sideslip angle can be measured from the phase of this signal. The evaluation of the amplitude and phase of the measurement signal is performed in a computing device. For direct measurement methods, LASSIE−System
(Low AirSpeed Sensing Indication
Equipment), i.e. 0°α, mounted on a bracket on the side of the cell below the main rotor
Also included are Cardan's suspended pitot tube measuring instruments and relative wind angle sensors with a permissible attack angle range of 360° and a sideslip angle range of −60° β60°. In LASSIE, it is possible to distinguish between operating areas slow flight and high speed flight. In slow flight the sensor is in the wash of the rotor. There, the relative wind velocity is superimposed with a very high wash velocity of the rotor, so that the dynamic pressure is satisfactory and analyzable. Using the relative wind angles α, β and the corresponding trigonometric formulas, the longitudinal velocity component u _A and the transverse velocity component v _A are calculated. Since in high-speed flight the sensor is outside the rotor wash, it is additionally possible to measure the vertical relative wind component w _A . In high-speed flight, the measuring principle of the LASSIE device is directly comparable to the corresponding sensor in winged vehicles, but in slow-speed flight it consists of a combination of true wind (Flugwind) and rotor wash. Flight speed is estimated from relative wind. Studies have shown that LASSIE devices exhibit varying degrees of accuracy depending on whether the sensor is in or outside the rotor wash. When sway and vertical flight conditions are also considered, the transition region where the LASSIE sensor is at the edge of the rotor vortex and sends out erroneous results is very large. Due to the very complex airflow conditions at the location of the measurement sensor,
The LASSIE device will require extensive calibration. The calibration results can then be used to compensate for errors that depend on the flight conditions. Both LORAS and LASSIE devices are available in mass production and are used on a variety of different types of helicopters. In addition to the current lack of accuracy, both devices have the disadvantages of being heavy, complex, and expensive. Furthermore, the equipment is very susceptible to damage and redundancy is costly to implement. As an alternative to the above devices, methods have recently been developed for calculating flight speed from the following measured quantities: maneuvering position, attitude angle, and rotational speed. i.e. VIMI-Method (Vitesse Indiquee par Moyen
Internes, FAULKNER and LAASH−Litef
Analytical Airspeed System for Helicopters)
This method is based on The basic advantage of these devices is that they are less susceptible to damage. The VIMI method is based on helicopter balancing, which represents the balance of gravity, drag and thrust. In order to determine the flight speed component, the balance adjustment equation is significantly simplified and modified: the longitudinal speed component u _A is then:
The lateral velocity component v A can be expressed as a function of the lateral steering position δ _Q and the pitch angle φ, and the lateral velocity component v _A can be expressed as a function of the lateral steering position δ _Q and the pitch angle φ. (1) u _A = h ₁ δ _L + h ₂ 〓 (2) v _A = h ₃ δ _Q + h ₄ 〓 In the improved Super-VIMI, the following quantity is additionally calculated in the two equations, i.e., the simultaneous maneuvering position. The dependence on δ _K , flight mass m and air density ρ is taken into account. Details regarding these additional terms are not publicly known. According to the company's data, the accuracy achieved is truly impressive. Although the available instructions do not clarify to what extent this method is applicable in unsteady flight conditions and vertical flight,
It is likely that the accuracy will be reduced. Furthermore, it is expected that the results will be significantly degraded by input signals that are prone to instability due to the large adjustment movements that the pilot must make, for example to adjust for unstable helicopter motion. In order to improve the "selectivity" of the corresponding measured variable, it is proposed to filter the steering input signal. However, in most cases, filtering results in a significant deterioration of the temporal characteristics. Studies have further shown that calculating longitudinal and lateral velocities (VIMI) separately is insufficient to obtain satisfactory accuracy for longitudinal and lateral velocities over a relatively large flight speed range. Because they are interconnected, they are known to be insufficient. The starting point of FAULKNER's method is that flight speed can be effectively estimated from rotor system reactions in low speed ranges. The basis of this method is the differential equation of the rotor blade runout motion. A quasi-stationary differential equation of runout motion is solved and is solved, ie, inverted, depending on the longitudinal and transverse velocity components. The input quantities to this highly sophisticated system are periodic and simultaneous maneuver position, rotational speed and rotor mast moment. Using this, the deflection coefficient is calculated. This method requires complete knowledge of the flight vehicle and aerodynamic conditions of the helicopter rotor. To arrive at a solution, additional empirical assumptions are needed to calculate the air flowing at the rotor. The measurement records of flight experiments using the “FAULKNER” method have given satisfactory results for the longitudinal velocity profile. The steady-state accuracy in transverse velocity is not fully satisfactory. There is no indication to what extent this method can be used for ascending and descending flights. Furthermore, there is the fundamental problem that the inversion of balanced systems is very sensitive to noisy input signals. In this case, noise is also a high frequency adjustment signal fluctuation. In the measurement records this was then significantly noisy in comparison with the reference velocity from the Doppler system,
Shown at estimated flight speed. The LAASH method is only effective in slow horizontal flight. The reason for this is that if the flight mass is known, only under this precondition can a unique relationship between the simultaneous maneuvering position and the absolute value of the flight speed hold. Depending on the absolute value of the flight speed, if the roll angle is known, a characteristic maneuver profile occurs with respect to the roll angle. In the LAASH-method, the absolute value of the flight speed, determined from the simultaneous maneuvering position signal, is used to select from a data file the maneuvering characteristics (correspondence of periodic maneuvering position to sideslip angle) corresponding to this speed, and the sideslip Angle is required. Ambiguities exist both in the longitudinal and lateral maneuver profiles, ie different values of the skid angle can occur. Due to the phase shift of both maneuver profiles, the corresponding sideslip angle of one can be estimated by comparison. Hovering area and transition area (25m/sV
Since the simultaneous maneuvering position at 35 m/s) has only a slight slope with respect to the flight speed, it is already prone to errors when determining the absolute value of the flight speed. Furthermore, just in the hovering region, the response of the periodic maneuvering position to the sideslip angle is strongly dependent on the flight speed and is therefore extremely sensitive overall to measurement errors and parameter changes. In this method, it is disadvantageous that there is ambiguity in the simultaneous maneuvering positions with respect to the flight speed. For example, the same simultaneous maneuvering position is required at hovering and at a flight speed of about 45 m/s. LAASH does not take this ambiguity into account, so it is necessary to estimate the speed range in advance. LAASH thus has two strict conditions that severely limit its practical use. Therefore, in the indirect (analytical) measurement method, the flight speed is determined by making full use of flight mechanical relationships. All methods utilize only stationary to quasi-stationary relationships. After all, all methods are based on imperfect flying machine models, and therefore produce more or less significant errors depending on the actual flight operation. In summary, the following can be confirmed for the device and method described so far: The entire sensor device is directly influenced by the rotor airflow conditions. Since the rotor airflow conditions are highly dependent on the flight operation, this has a detrimental effect on the accuracy of the sensor arrangement and currently requires compensation for errors that are dependent on the flight conditions. Analytical methods take advantage of the fact that in slow flight, maneuver changes are large but speed and direction changes are small. An advantage of the analytical method is that all the necessary sensors are housed within the room and are therefore less susceptible to damage. Essentially analytical methods have advantages over equipment-technical solutions. However, the known analytical methods as a whole have limitations in their use:
The LAASH and "FAULKNER" methods are only effective in slow flight and hovering, and the VIMI method also has a non-linear course of maneuver amplitude and attitude angle, although it does not completely negate this method. It is expected that accuracy regarding flight speed will be significantly impaired. Another disadvantage is that errors can be even greater due to dynamic flight operations. In both methods, the vertical velocity component (or overall angle of attack) is measured. Therefore, the flight speed can only be determined incompletely. The object of the invention is to provide a flight speed measuring device which is effective over the entire flight speed range of a helicopter, applicable both in static flight conditions and in dynamic flight operations. Means for Solving the Problem and Effects of the Invention According to the present invention, this problem is solved by the configuration described in the characterizing part of claim 1. Advantageous embodiments of the invention are described in the subclaims. In the present invention, the flight speed is measured using an interlocking measuring device that processes the maneuvering position, attitude angle and rotational speed. If it also measures the path velocity and is used in a suitable connection, it can increase the accuracy during flight operations and measure the wind speed. The following measured quantities: air temperature T _s and static pressure p _s
can be used to match atmospheric conditions. In the method for measuring flight speed, measurements of the maneuvering position, pitch and roll angles and rotational speed are provided to a computing device. Using the calculated flight speeds, the method was adjusted to the actual flight conditions at the time (fast flight, slow flight, lateral flight and vertical flight) and three flights were performed using the above measurements. Velocity components (longitudinal, lateral, vertical) are calculated. The advantages of the device of the invention compared to existing solutions include the following: - All three flight speed components are measured, namely the absolute value of the flight speed and the angle of attack and sideslip angle. - This measurement method works accurately even in dynamic flight operations and only delivers average values. It's not about doing it. - Based on the clearly demarcated internal structure of the method, adjustment to changes caused by atmospheric changes or shape can be performed easily and accurately; - The calculation method can be matched to the flight conditions at the time. This allows the given conditions of the helicopter, which vary significantly depending on the flight conditions, to be fully taken into account. Embodiments Next, the present invention will be explained in detail with reference to the drawings, with reference to the illustrated embodiments. According to LUENBERGER's theorem, it is well known that an unknown (unmeasurable) state quantity can be estimated based on a known (measurable) quantity by means of a state observer: the state vector x and A dynamic process with a model matrix A of system characteristics, a vector u of input quantities, and a model matrix B of input characteristics (3) x = A x + B u and a measurable output vector (4) y = C x A model is developed on this basis and operated with the same input quantities u as the process itself. Difference between measured and estimated output
With a feedback of y = y - y and an amplification with K , the transfer characteristics of the observer can be matched to the specific setup task. In this way, the differential equation (DGL) of the observer is obtained: (5 ) x= A x^+ Bu + KC ( x - x^) Figure 1 shows the dynamic process using the LUENBERGER observer. The block diagram of From the difference between equations (3) and (5), a differential equation for the estimation error can be obtained: (6) x = ( A - KC ) is a zero vector, ie, even for initial values that are not identical to the process, the observer generates a complete, error-free state vector after decay of the initial error. The basis of the observer is therefore the mathematical modeling of the process characteristics. The core of the invention is a simple dynamic model of helicopter motion that can be easily matched to the given physical conditions of the helicopter, and that allows observers to appropriately capture the characteristics of the helicopter that change with flight conditions. This is something that can now be taken into account. For this task, only the motion of the helicopter center of gravity relative to its surroundings is important, so the differential equations of motion of a rigid helicopter are considered as the basis for modeling. The following equation holds from the force and moment balance conditions expressed in the coordinate system fixed to the helicopter: Differential equation (7) represents the equilibrium position of inertial force, dynamic aerodynamic force, gravity and their moments. The inertial force and the moment of inertia are the inertial state quantity V _K ,
Ω and the weight depends on the inertial attitude angles φ, θ, the dynamic aerodynamic forces and their moments result from the relative motion of the aircraft with respect to the surrounding atmosphere. Based on the assumption that the dimensions of the aircraft are small compared to the shortest wavelength of air motion, the following equation holds. (8) V _K = V + V _W This equation represents that the path speed V _K is vector-synthesized from the flight speed V and the wind speed V _W. Similarly, the following equation holds for rotational speed: (9) Ω = Ω _A + Ω _W A perfect approximation of the dynamic aerodynamic forces and moments, which are only generally formulated in equation (7). The calculations are very complicated and cannot be practically used for a setting problem such as this one. If we are only interested in the motion of an idealized helicopter as a rigid body, series expansion is also sufficient for state and adjustment variables to represent dynamic aerodynamic forces and their moments. For _example , ^the following _equation holds for the longitudinal _{force : ( 10} ) For example, it is expressed as follows. (11) X _u = _{c Xu} q _{Rp S} where c _Xu denotes the dimensionless longitudinal force derivative dc In order to obtain a valid value even during hovering, the dynamic pressure q _Rp is determined by the rotor circumferential speed U _Rp in a helicopter.
Related to: (12) q _Rp = ρ _s /2U ² _Rp The air density ρ _s is the next measurable quantity, i.e. the static pressure p _s
and can be expressed as a function of the air temperature T _s and the specific gas constant R of the air: (13) ρ _s = p _s / _RT sCurrently, the operating characteristics of helicopters vary significantly with flight speed. Therefore, equations (9), (10),
(11) and constant derivatives cannot satisfactorily represent dynamic aerodynamic forces and moments over relatively large flight regions. In order to model the operating characteristics of a helicopter over a relatively large flight range, linear derivative models are used for different flight speeds and interpolation between the individual derivatives is performed. In other words, it is continuously switched between models. Therefore, the dynamic aerodynamic forces and moments are calculated according to the following formulas: In this case, the following formula holds for the aerodynamic system matrix and the input matrix: (15) A * = A * _O +u _A A * ₁ +u ² _A A * 2 + u ³ _A A * 3 +... (16) B * = B * 0 + u _A B * 1 + u ² _A B * 2 + u ³ _A B * 3 +... The derivative with respect to the reference state can be determined by the helicopter manufacturer. What degree of polynomial is useful in equations (15, 16) depends on the given conditions of each helicopter. For the interpolation in equations (15, 16), it is generally sufficient to use the cube of the flight speed. It is true that when air flows through the rotor, the flight speed is superimposed on the local rotor circumferential speed, so dynamic pressure dependence can already be taken into account using the square term, but even more expansion terms It is effective if you use
In this case, it is recognized that the cube term as shown in equations (15, 16) is extremely advantageous. Extreme deviations from vertical flight conditions do not occur, so
There is no need to match the aerodynamic system and input matrix to eg lateral or vertical velocity components. The reason for this is that the first-order term is sufficiently accurate in the relatively close vicinity of the operating point at that time. If even large deviations from pure longitudinal motion are to be detected with satisfactory accuracy, Equation (15,
16), it is also conceivable that a transverse velocity component or a vertical velocity component and a combination of all three components are used for updating the dynamic air matrix. Because of the relationship shown in equations (11, 12), it is necessary to consider the atmospheric conditions at that time. That is because aerodynamic forces and moments depend on the air density ρ _s . Since the derivatives are related to the standard atmosphere, the “dynamic aerodynamic matrix” A * and B * in equation (15) should be further multiplied by the coefficient ρ _s /ρ _o , resulting in (13) The following formula holds true using: (17) A ′=(P _s /P _o )(T _o /T _s ) A * (18) B ′=(P _s /P _o )(T _o /T _s ) B * This system matrix and input matrix are the relative wind components (Anstro¨mkomponent) u^ _A , v^ that exist not as directly measured quantities but as calculated values in equations (15, 16). _A , updated via w^ _A . In this way, this equation differs from traditional equations in which measured quantities are utilized for model matching. Also, path velocity vector V _K = [u _K , v _K , w _K ] ^T
If, for example, can be used as the output of a Doppler navigation system, the accuracy in curved flight can be improved as follows. That is, instead of the estimated flight speed V , the measured path speed V _K is used to determine the inertial forces. Figures 2 to 5 show the principle of operation of the system for determining the route. Using the circuit shown in the block diagram, (15,
16), equation (7) is simulated. Via a correction element, this system is adapted to the flight conditions of the helicopter. Since theoretically processed models often do not correspond exactly to given conditions in reality, the method will be described further below. That is, if you use this method,
It is possible to increase the steady-state accuracy of the system for determining the route and furthermore reduce the computational cost of the proposed method. In equilibrium (steady) flight conditions, no acceleration occurs, so the following equation holds: (19) x = A x _Tr + Bu Tr _! = O If the observation device is to operate regularly without any estimation errors, the amount of acceleration of the observation device must also be eliminated. Therefore, since x _Tr −x^ _Tr = O , the following equation arises: (20) A^=x^B^ _Tr +x^ u _Tr ! = O If there is a model error, i.e. A^ = A + Δ A and B^ = B + Δ B , (20) is not satisfied. If the observer is nevertheless to operate regularly and without estimation errors, compensation for model errors is necessary, which can be done as follows. That is, by substituting the states and input quantities valid for the equilibrium flight state, which can be measured relatively easily, into the differential equation (20), the acceleration quantity that disturbs the setting of the steady state of the observer can be obtained. : (21) f = x = Δ A x s _Tr + Δ Bu _Tr When this acceleration is subtracted in the differential equation (20), the observer can also be modeled with errors for the equilibrium flight state. Regardless of the situation, it is guaranteed to operate without any estimation error: (22) x = A^x^ + B^ u + KC ( x −x^) − f (x^) Usually, a correction term f (x^) is not constant for all states. However, since the "observer error" can be expressed as a function of an appropriate state quantity from equation (21), a perfect approximation can be realized using polynomials corresponding to equations (15, 16). . FIG. 2 schematically shows a helicopter 1 with a steerable main rotor 2 and a tail rotor 3. FIG. The helicopter undergoes periodic main rotor blade adjustment δ _L ,
A steering mechanism 4 is provided for δ _Q and simultaneous main rotor blade adjustment δ _K and tail rotor adjustment δ _S. The steering position is determined by a suitable sensor (distance or angle detector), shown here as block 5.
Measured by The measured values are supplied to the signal processing unit via line 6 as analog or digital electrical or optical signals. Measuring and transmitting the lateral steering position can optionally be omitted. Depending on the measured steering position, the rotor control (swath plate, “Spinne”)
The important point is that it enables a unique correspondence to the position of . If a regulator is used, measurements must be taken after each regulator operation. Similarly, the helicopter's rotational speed p around the vertical axis, rotational speed q around the horizontal axis, rotational speed r around the vertical axis, roll angle φ and pitch angle θ as attitude angles are measured. be done. In order to increase the precision during flight operations (for example curved flights) using the rotational speed, it is advantageous to measure the path velocity component u _K , the transverse component v _K and the vertical component w _K in the longitudinal direction of the helicopter. However, this is not absolutely necessary for the basic functionality of the method. Sensors for rotational speed and attitude angle are shown as block 7. Its output, after being transformed in a suitable manner, is supplied via line 8 to a signal processing section. In order to match the atmospheric conditions and thus to increase the accuracy, sensors 9 for static pressure P _s and ambient temperature T _s are provided. Its output is modified in accordance with the control signal and transmitted via line 10. FIG. 3 shows a first embodiment of a signal processing circuit with input lines 6, 8 and 10 for measurement data of sensors 5, 7 and 9. FIG. The circuit shows a measuring device 11 comprising two modules, each illustrated by a block surrounded by a dashed line. Block 12 is a module for input characteristics of aircraft motion. Blocks 13 and 14 represent modules of the system properties of the vehicle. Block 12 receives cyclic and simultaneous control signals (Zyklische und Kollektive) as input variables.
Steuersignal), optionally a signal for the lateral steering position, and additionally signals for the air pressure P _s and the ambient temperature T _s . Block 12 has amplification matrix circuits B _O to B ₃ to which input signals are supplied. In these matrix circuits these signals are processed according to the following formula, which applies to each of the six outputs x _Bij of the four matrix circuits. (23) x _Bij =B _ij 1δ _L +B _ij2 δ _Q +B _ij3 δ _K +B _ij4 δ _S (i=0...3) (j=1...6) Here, Bijk refers to the value used in memory. Derivatives provided by the manufacturer of the input characteristics, stored so that
It is. The amplification matrix circuits B _O to B ₃ each have a different derivative, depending on the flight state. The outputs of the matrix circuits B _O to B ₃ are multiplied together using multipliers 12.5 according to equation 16 and deliver the output of this model as output vector x _B . The individual vector elements are then respectively multiplied by the coefficients Ps/Pn and Tn/Ts, from which _the output vector x'B results. Corresponding to what has been explained for the model in block 12, in the model of the aerodynamic system properties in block 13 the estimated state quantities u^A, v^a and w^A of the translational velocity and the rotational velocity are p^,
q^ and r^ are multiplied by the amplification matrix circuit A ₀ to A ₃ and the multiplier 13.5 and the coefficients Ps/Pn and Tn/Ts provided and stored by the manufacturer. This results in the output vector x'A _of block 13. Block 14 of the model of system properties
In addition to this, an element that takes weight into account (element 1
4d), an element (element 14c) that takes into account inertial force or moment of inertia, and an element (element 14e) that takes into account kinematic optical relationships are provided. Furthermore, in block 14, an element is provided which is used to measure the inertial acceleration forces and to which an estimated quantity (position 14a) or a measured quantity (position 14b) can be applied. The equation shown in FIG. 3 is calculated for elements 14b, c, d and e. The output of block 14 is added to the output of block 13, except for the kinematically related output of block 14e. All outputs of blocks 13 and 14 are combined into an output vector x' _S of system characteristics. The measured outputs p, q, and r of the rotational speed and the attitude angles φ and θ applied to the input side 8 are determined by the state quantities p^,
q^ and r^ and Φ^ and Φ^ are compared and the difference is fed to a correction matrix circuit 18 where a combination is carried out according to equation (23). From this estimated state vector x^, the estimated state quantities of the three translational velocities u^ _A , v^ _A and w^ _A are transferred to the indicating device 19. The circuit of FIG. 4 largely corresponds to the circuit of FIG. The difference is block 13 in Figure 3.
to the input side of the block 23 corresponding to the block 24 corresponding to the block 14 in FIG.
The measured quantities are applied for rotational speed and attitude angle. Otherwise, the output vectors x' _B and x ' _S are obtained in the same way and are added to the adder 2.
5 is added. The input 8 for rotational speed and attitude angle is here connected to a first correction element 28 in which calculations according to equation (23) are performed. Correction element 2
The first three outputs of 8 are summed with the three outputs of the integrator 27 in an adder 29, resulting in estimated flight velocity components v^ _A , v^ _A and w^ _A , which are are fed to the blocks 23 and 24 of the model for the system properties and at the same time to the indicating device 30. The output of the adder 25 and additionally the output variable for the kinematic relationship in the output vector x' _S is fed to a second correction element 26, where an operation corresponding to equation (23) is carried out. Correction element 2
The three outputs of 6 are fed to an integrator 27. The circuit of FIG. 5 corresponds in its basic configuration to the circuit of FIG. 3. Only a portion is shown here. The circuit of FIG. 5 has additional circuits 31/32 for compensating for stationary model errors. This circuit 31/32 is additionally supplied with the estimated values for the flight speed components u^ _A and v^ _A as input variables. Output x to two branch circuits 31/32
_KU and x _KV occur according to the equation: That is, (24) x _KU = u^ _A f _u1 + u^ _A ² f ² + u^ _A ³ f _u3 This equation also holds true for x _Kv . A further constant vector f _O is added to the power quantities x _Ku and x _Kv , which are dependent on the estimated flight speed components u^ _A and v^ _A. The total sum x _K is additionally supplied to adder 15 . The compensation circuit of FIG. 5 can also be provided in the circuit of FIG. 4. The output vector x _K is in this case supplied to the adder 25 . The compensation vectors f _o , f _ui and f _vi are expressed by Equation (14)
It is obtained according to (22). In circuits 31 and 32, the estimated flight speed components u^ _A to v^ _A
Three multiplications are performed. It is also possible to use polynomials of lower or higher order. This also applies to the multiplication with the estimated flight speed component u^ _A in the circuits of FIGS. 3 and 4. In FIG. 5, a matrix circuit C35 is provided to which the output of the integrator 16 is applied. This matrix circuit takes into account the fact that not all state quantities are necessarily measurable, and therefore only the corresponding quantities for comparison are available at the point of comparison between the estimated and measured output quantities y^ and y. ing.
The vector of estimation error y^ is derived from the differences p^, q^, r^ between the estimated and measured rotational speeds and the differences Φ^ and Θ^ between the measured and estimated attitude angles. It is completed. Furthermore, instead of the velocity quantity r^ _A , depending on the respective given conditions, a velocity quantity v^ _A , but possibly a measured quantity to match the current flight conditions, can be substituted in block 12. It can also be used for multiplication in and 13 to 22, 23. When using the compensation circuit shown in Fig. 5, blocks 12, 13 to 22, ₂ can be
3 to the flight condition may also be omitted in some cases. Instead of the three flight speed components, the absolute value of the flight speed and the absolute values of the angle of attack and sideslip angle may be determined. In a similar manner to matching for translational flight conditions, matching for curved flight can also be achieved by appropriate use of rotational speed or attitude angle, with the resulting improved accuracy being measured. Become. The fact that data regarding the mass and center of gravity of the helicopter must be transmitted in advance to the device for measuring the flight speed is certainly also disadvantageous. For example, the pilot can enter corresponding values into the calculation unit via a keyboard device. However, the mass of a helicopter can also be determined as follows. That is, erroneous flight mass data will result in erroneous estimates of vertical velocity, so if a vertical velocity meter (barometer type vertical velocity) is present, the measured vertical velocity and the calculated vertical velocity used in observation instruments through comparison with Shifting adjustments to the value of the vehicle mass can be made until the vertical velocity error is eliminated. In this way not only the modeling in the observer is corrected, but also the calculated vehicle mass can be used by the pilot, for example, for power calculations. Finally, the meanings of symbols used in this specification will be explained. A- based matrix. A ★ System matrix of dynamic aerodynamic forces and their moments. A ★★ System matrix of dynamic aerodynamic forces and their moments normalized to mass and moment of inertia. Δ Model error of A -system matrix. B input matrix. B ★ Input matrix of dynamic aerodynamic forces and their moments. B ★★ Input matrix of dynamic aerodynamic forces and their moments normalized to mass and moment of inertia. Δ B Model error of input matrix. B _ijk Coefficients of input matrix. i=1...3,j
=1...6, k=1...4. C output-(measurement-) matrix. I unit matrix. I moment of inertia. K correction matrix. L ^A Aerodynamic swaying moment. M ^A Aerodynamic pitching moment. ^N AAerodynamic sway moment. S Standard area. T Reference temperature = 288.15K. T Ambient temperature. U _RO rotor peripheral speed. V flight velocity vector = V = [u _A , v _A , w _A ] ^T. V _K path velocity vector = V _K = [u _K , v _K , w _K ]
^T. V _W wind speed vector V _K = [u _w , v _w , w _w ] ^T. V Absolute value of flight speed. X ^A Aerodynamic longitudinal force. Y ^A Aerodynamic lateral force. Z ^A Aerodynamic normal force. f _O correction vector. f _iu Correction vector for longitudinal velocity component. f _iv Correction vector for the lateral velocity component. h _i Amplification factor for VIMI. m mass. P _o reference pressure = 1013.25hpa. P _s static pressure. P: Lateral angular velocity. q pitch angular velocity. q Dynamic pressure at the tip of _{the RO} rotor blade. r One-way angular velocity. u input vector u = [δ _L , δ _Q , δ _K δ _S ] ^T . u Longitudinal velocity. v Lateral velocity. w Vertical velocity. x state vector x = [u _A , v _A , w _A , p, q,
r, φ, θ〕 ^T . x Vertical axis. y output vector y = [p, q, r, φ, θ] ^T. y horizontal axis. z Vertical axis. α Angle of attack. β Side slip angle. δ _K simultaneous control position. _δ LLongitudinal steering position Lateral steering position Synchronous steering position. δ _S lateral steering position (tail rotor). ρn Air density when p=1013.25hpa. T=288.15K. ρs Air density. φ Rolling angle. θ pitch angle. Ω rotation vector Ω = [p, q, r] ^T. ∧ Estimated amount. ~ The difference between the measured quantity and the estimated quantity.

[Brief explanation of the drawing]

Fig. 1 is a block diagram illustrating the dynamic process using the observation instrument by LUENBERGER, Fig. 2 is a schematic diagram of the entire helicopter including the control mechanism and sensors, and Fig. 3 is a block diagram showing a first embodiment of the signal processing circuit, FIG. 4 is a block diagram showing another embodiment of the circuit in FIG. 3, and FIG. 5 is a block diagram showing another embodiment of the circuit in FIG. FIG. 3 is a block diagram showing another embodiment of the circuit. 11, 21... Measuring device, 12, 22... Model of input characteristics, 13, 14, 23, 24... Model of system characteristics, 18, 26, 28... Correction element, 1
6, 29... Integrating stage, 19, 30... Indicating device,
31, 32...Model error compensation circuit.

Claims (1)

[Claims] 1. An apparatus for measuring the flight speed of a helicopter using periodic and simultaneous control signals and attitude angles for roll and pitch, comprising: periodic and simultaneous control signals δ _L , δ _Q , δ _K and attitude angles φ, θ and helicopter axes x, y,
Model 1 of the input characteristics of the helicopter motion, in which rotational speeds p, q, r about z are supplied and two models, namely periodic and simultaneous maneuver signals 6, act as input quantities.
2, 22, and models 13/14, 23/24 of helicopter system characteristics in which the state quantities act as input quantities; and a correction device for the measured quantities of rotational speed and attitude angle. 18, 26/28, and the estimated state quantity x^ is the same as the above two models 1
Output x′ _B of 2, 13/14, 22, 23/24,
x′ _S and an integration stage 16 derived by integrating the algebraic sum of the output signals of the correction devices 18, 26.
and after the integration stage 16, 29 the value is taken out,
A flight speed measuring device characterized in that it is equipped with an indicating device 19, 30 for three estimated speed components u^A, u^A, w^A. 2 Two models 12-13/14, 22-2
2. A flight speed measuring device according to claim 1, wherein 3/24 can be matched to the current flight state via estimated state variables. 3. The flight speed measuring device according to claim 2, which is provided with at least one circuit for matching the two models 12-13, 22-23 to the current flight conditions in proportion to their speeds. . 4. Claim 3, which is provided with a plurality of parallel circuits that perform second-order and higher-order polynomial operations.
Flight speed measuring device as described in . 5 Models 13, 14, 23, 24 of system characteristics are dynamic aerodynamic forces, their moments and weights 14
d, 24d, inertia force or moment 14c, 24
c and/or kinematic relationships 14c, 14
2. The flight speed measuring device according to claim 1, further comprising an element that takes into account e. 6 The model of system characteristics is based on the estimated state quantities14
a, 24a and/or the measured state quantity 14
Claim 1 includes an element for measuring an inertial acceleration force to which b, 24b can be applied.
Flight speed measuring device as described in . 7 The measured quantities of rotational speed and attitude angle are applied to the correction devices 18, 28 and the adder 1
5, 25, the output sides of the two models x ′ _B , x ′ _S
2. A flight speed measuring device according to claim 1, wherein: is connected in advance. 8. The flight speed measuring device according to claim 7, wherein the difference between the estimated state quantity and the measured state quantity is applied to the correction device 18. 9. The correction device has a first correction element 26 between the addition section 25 and the integration stage 27 and a second correction element 28 on the input side of the measured state quantities for the rotational speed and attitude angle. A flight speed measuring device according to scope 1. 10 To compensate for stationary model errors, 2
A circuit 31, 32 is provided which is connected to the summing unit 15 between the two models 12, 13/14, into which at least one estimated or measured component of the flight speed u^ _A , u ^ The flight speed measuring device according to claim 1, wherein _A is supplied as an input. 11 To compensate for stationary model errors, 2
A circuit 31, 32 is provided which is connected to the adder 15 between the two models 12, 13/14, into which at least the measured or estimated quantity of the attitude angle or rotational speed component is input. A flight speed measuring device according to claim 1, which acts as a flight speed measuring device. 12 The circuit has at least two parallel branches 31,3
Claim 10 or 11 having 2.
Flight speed measuring device as described in . 13. The flight speed measuring device according to any one of claims 10 to 12, wherein the circuit has a plurality of parallel circuit branches for calculating second-order and higher-order polynomials. 14 Two models 12, 13/14, 22, 2
In addition to 3,24, the static pressure p _s and the ambient temperature T _s
2. A flight speed measuring device according to claim 1, wherein the signal is supplied as a correction coefficient. 15. Claims 1 to 14 in which, if the barometric or inertial vertical velocity component is known, a correction factor for the assumed helicopter mass is determined by comparing it with the estimated vertical velocity component. The flight speed measuring device according to any one of the preceding items.